Raising: jai


This section is long and complicated. On the plus side, it's also the final section in the course.

Things weren't always like this. In the '80s, the ancestor of Lojban still said that things were interesting, and people were crazy, just like most normal languages, and without detouring through abstractions. So what happened?

Well, what happened was that Lojbanists noticed how linguists have been analysing these concepts in natural languages, and how they were coming up with their own versions of selbri. Often, what was a noun in one part of the sentence, and a verb in another part, were brought together and considered to be underlyingly part of the same abstraction sumti.

Note: The word for selbri in English, by the way, is predicates; we've been avoiding it up to now, but we think you can handle the truth from now on...

A good example is the phrase I am difficult to annoy in English. At first sight, you might think that I is a sumti of difficult. And grammatically it is: it's the subject. But logically it isn't: what we're describing as difficult is not me. We can't say:

What's actually going on is that, underlyingly, what is difficult is to annoy me: the action of getting me annoyed is what is hard to achieve — not me! This is why English also allows you to say It is difficult to annoy me, and (if you squint a little) To annoy me is difficult. And sure enough, Lojban expresses this concept according to that 'underlying' form:

lenu fanza mi cu nandu
The event of annoying me is difficult

So why did English pull that weird switcheroo with I am difficult to annoy? Basically, because when we talk, we aren't concentrating in our minds on intangible abstractions like "the event of annoying me", let alone "the state of Jyoti having certain unspecified properties." Instead, we run little stories in our head, with heroes and villains: concrete heroes and villains — people, for the most part. And as it happens, we make the subjects of our sentences be the heroes and villains we're concentrating on. (That's what a subject's ultimate job is: to present what we're concentrating on.)

So by pulling a switcheroo like that, we're not talking about abstractions and events any more; the subject of the sentence is now our perennially favourite subject — namely me: it's me that is difficult to annoy. (Yes, it is all about me...) This process is called in linguistics raising, because it raises concrete subjects (and objects) we want to talk about, out of the haziness of an abstraction sumti (or 'clausal argument', to use English logical terminology.)

Once the requisite number of Lojbanists did an undergraduate course in syntax (you may commence throwing darts at effigies of Nick Nicholas at your leisure), it was realised that there were a lot of gismu whose place structures contained both a raised concrete sumti (usually x1), and an abstraction sumti which itself contained the first sumti. For example, the place structure of fenki used to be

x1 is crazy in behaviour x2 (abstraction) by standard x3

But any abstraction that would go into x2 would contain the x1 sumti: any crazy behaviour would automatically be the behaviour of the crazy person. For example, you'd get

The question then became: does the x1 tell us anything the x2 wasn't already telling us? We know who was involved in the crazy behaviour, because that person would be a sumti inside x2. (More specifically, he or she would be the active party: someone hitting random strangers is crazy; someone being hit by random strangers isn't — although arguably someone allowing themselves to keep being hit by random strangers is.) Was there any reason, then, to grant the person an extra place in the overall bridi? The decision was, no: behaviour is what is crazy, so you can work out that the person acting out the behaviour is the crazy person. There's no need to have an extra place for the person, when you can already work out who they are. The same conclusion was arrived at for cinri: it is abstractions — events and qualities — that attract interest; and an interesting person is simply a person involved in an interesting abstraction.

All well and good; but natural languages do raising for a reason. So when Lojban has its gismu without raising, it gains in eliminating redundancy and logical muddledness; but it loses in 'naturalness'. We like talking about people rather than abstractions in our languages; and Lojban should not go out of its way to form an exception to this.

There is a solution of sorts to this problem using tu'a; but it doesn't actually do what raising does in natural languages: it doesn't change the x1 place from an abstraction to a concrete sumti. And there are times you will want to do just that.

One example is joining bridi-tails. In English, you can say Jyoti is interesting and beautiful. This is based on two sentences (Jyoti is interesting, Jyoti is beautiful) which have the same subject. So we can easily combine them into a single sentence. In Lojban, the equivalent sentences are

tu'a la djiotis. cinri


la djiotis. melbi

There is no way you're going to join those two bridi together with gi'e: they simply do not have their first sumti in common. But they're both somehow 'about' Jyoti; so you really should be able to work around this.

An even more important instance when you want raising is in forming sumti out of this kind of gismu. A sumti means whatever goes into the x1 of its selbri. If la djiotis. ninmu "Jyoti is a woman", then I can describe Jyoti as lo ninmu 'a woman'. If lemi karce cu xe klama le gusta fu mi "My car is a vehicle to the restaurant for me", then I can describe lemi karce as lo xe klama 'a vehicle'. So how do I say that someone is a cheat, or a deceiver? The gismu for 'deceive', tcica, has the place structure

x1 (event/experience) misleads/deceives/dupes/fools/cheats/tricks x2 into x3 (event/state)

This means that, while in English we say that "x1 (person) deceives x2 into doing x3, by doing x4", in Lojban the person and the action are merged into the one place. That makes lo tcica a trick, not a trickster; a deception, and not a deceiver. To say that someone is a trickster or a deceiver, we need to use tu'a: tu'a da tcica. But you can't put lo in front of tu'a da: the deceiver has to be the x1 of some selbri, in order to get their own sumti.

The solution to this is to force Lojban to have raising after all, changing the place structure of the selbri involved. This works just like se changing the place structure of its selbri, swapping its first and second place. If we put jai in front of a selbri, its x1 place changes from an abstraction, to any sumti contained within the abstraction. Let's try this with a few sentences:

You'll notice that, with these new place structures, the Lojban phrases sound pretty much like their English equivalents. For example,

la djiotis. jai cinri 
Jyoti is interesting

la ranjit. jai tcica la suzyn.
Ranjeet deceives Susan

We can now do with jai those things we couldn't before. The Lojban for "Jyoti is interesting and beautiful", for example, is

la djiotis. jai cinri gi'e melbi

That's because Jyoti goes in the x1 place of jai cinri, just as it goes into the x1 place of melbi. And if I want to make a sumti meaning 'deceiver' or 'trickster', I can use jai to do it:

tu'a la ranjit. tcicala ranjit. jai tcicalo jai tcica

However, mi jai nandu does not correspond to "I am difficult to annoy." In switching a concrete sumti for the original x1 — the abstraction that was difficult — we have lost the abstraction itself: there is nothing in mi jai nandu that means 'to annoy'. But not to worry: Lojban allows you to keep the original abstraction in the bridi by preceding it with fai. fai is a place tag like fa and fe; it effectively adds a new place to the bridi. So I am difficult to annoy is matched almost word-for-word by the Lojban sentence

mi jai nandu fai lenu fanza mi

And we can apply this pattern further afield; for example, "the book took three months to write" is in Lojban properly

lenu finti le cukta cu masti li ci 
To write the book had a month-duration of three

Raising allows the slightly more familiar-looking

le cukta cu jai masti li ci fai lenu finti

jai has not proven as popular as tu'a, presumably because it involves a fairly thorough rearrangement of place structures — and has the whiff about being somehow 'un-Lojbanic'. But as we've seen, it allows you to talk about things in a way that is in many ways more natural; and though it belongs to 'advanced' Lojban, it is a feature you will find it useful to be familiar with.

Exercise 4

That was pretty heavy going. You can relax: this exercise will go easy on you. (You still have the final translation exercises to go through, after all!) Where possible, and by all means necessary, recast the abstractions in the following sentences so that they use jai (and fai, where applicable.)

  1. .i tu'a mi nabmi

  2. .i ledu'u mi xebni loi kensa fange cu nabmi

  3. .i mi djuno tu'a la lojban.

  4. .i mi djuno ledu'u la lojban. cu bangu kei la lojban.

  5. .i lenu mi ckire da cu nibli lenu mi se xamgu tu'a da (Don't try and be too clever here — it won't work...)

  6. .i lenu lenu la jan. xalfekfri cu nabmi cu cizra (Only eliminate one level of abstraction.)

  7. .i da poi lenu fanza ke'a cu nandu cu zvati (Reduce this, then see if you can't reduce it a little more...)